Modeling tongue surface contours from Cine-MRI images.

Area of Science:

• Biomechanics
• Speech Science
• Medical Imaging

Background:

• Understanding tongue movement is crucial for speech production.
• Quantifying 2D tongue shape and position changes during speech is complex.
• Subphonemic speech events like coarticulation and asymmetries require detailed analysis.

Purpose of the Study:

• To develop a simple mechanical model for quantifying global tongue motion during speech.
• To represent 2D tongue surface shape and position differences during speech movements.
• To analyze speech events such as coarticulation and left-to-right motion asymmetries.

Main Methods:

• Utilized tagged Magnetic Resonance Imaging (MRI) to capture tongue motion during speech.
• Collected MR image-sequences during consonant-to-vowel (C-to-V) movements across three sagittal planes.
• Decomposed global tongue surface motion into translation, rotation, stretch, and shear using a mechanical model.

Main Results:

• The largest C-to-V shape deformation occurred from /k/ to /a/, involving vertical compression and horizontal expansion.
• Coarticulation involved a trade-off where tongue shape accommodation reduced movement distance.
• Left-to-right motion asymmetries potentially increased speech rate by reducing moved mass.

Conclusions:

• A simple mechanical model effectively quantifies tongue motion during speech.
• Tongue shape dynamics and asymmetries play significant roles in speech production and efficiency.
• Findings provide insights into articulatory phonetics and speech motor control.